Update app.py

app.py

@@ -10,10 +10,16 @@ from keras.layers import BatchNormalization, DepthwiseConv2D, TFSMLayer
 
 # --- Fix deserialization issues ---
 original_bn = BatchNormalization.from_config
-BatchNormalization.from_config = classmethod(lambda cls, config, *a, **k: original_bn(config if not isinstance(config.get("axis"), list) else {**config, "axis": config["axis"][0]}, *a, **k))
+BatchNormalization.from_config = classmethod(
+    lambda cls, config, *a, **k: original_bn(
+        config if not isinstance(config.get("axis"), list) else {**config, "axis": config["axis"][0]}, *a, **k
+    )
+)
 
 original_dw = DepthwiseConv2D.from_config
-DepthwiseConv2D.from_config = classmethod(lambda cls, config, *a, **k: original_dw({k: v for k, v in config.items() if k != "groups"}, *a, **k))
+DepthwiseConv2D.from_config = classmethod(
+    lambda cls, config, *a, **k: original_dw({k: v for k, v in config.items() if k != "groups"}, *a, **k)
+)
 
 # --- Constants ---
 IMG_SIZE = (224, 224)
@@ -52,8 +58,11 @@ def preprocess_with_steps(img):
     resized = cv2.resize(sharp, IMG_SIZE) / 255.0
 
     fig, axs = plt.subplots(1, 4, figsize=(20, 5))
-    for ax, image, title in zip(axs, [img, circ, clahe_img, resized],
-                                 ["Original", "Circular Crop", "CLAHE", "Sharpen + Resize"]):
+    for ax, image, title in zip(
+        axs,
+        [img, circ, clahe_img, resized],
+        ["Original", "Circular Crop", "CLAHE", "Sharpen + Resize"],
+    ):
         ax.imshow(image)
         ax.set_title(title)
         ax.axis("off")
@@ -74,9 +83,11 @@ def show_lime(img, model, pred_idx, pred_label):
             classifier_fn=lambda imgs: predict(imgs, model),
             top_labels=1,
             hide_color=0,
-            num_samples=100
+            num_samples=50,  # reduced samples for speed
+        )
+        temp, mask = explanation.get_image_and_mask(
+            label=pred_idx, positive_only=True, num_features=10, hide_rest=False
         )
-        temp, mask = explanation.get_image_and_mask(label=pred_idx, positive_only=True, num_features=10, hide_rest=False)
 
         fig, ax = plt.subplots(1, 1, figsize=(6, 5))
         ax.imshow(mark_boundaries(temp, mask))
@@ -91,24 +102,22 @@ st.title("🧠 Retina Disease Classifier with LIME Explanation")
 model = load_model()
 
 with st.sidebar:
-    uploaded_files = st.file_uploader("📂 Upload retinal images", type=["jpg", "jpeg", "png"], accept_multiple_files=True)
+    uploaded_files = st.file_uploader(
+        "📂 Upload retinal images", type=["jpg", "jpeg", "png"], accept_multiple_files=True
+    )
     selected_filename = None
     if uploaded_files:
         filenames = [f.name for f in uploaded_files]
-        selected_filename = st.selectbox("🎯 Select an image to preprocess and predict", filenames)
+        selected_filename = st.selectbox("🎯 Select an image to explain", filenames)
 
-# --- Process selected image ---
+# -- Predict & Display for Selected Image --
 if uploaded_files and selected_filename:
     file = next(f for f in uploaded_files if f.name == selected_filename)
-
-    # Read bytes once and reset pointer for later use
-    file_bytes = file.read()
     file.seek(0)
-
-    bgr = cv2.imdecode(np.frombuffer(file_bytes, np.uint8), cv2.IMREAD_COLOR)
+    bgr = cv2.imdecode(np.frombuffer(file.read(), np.uint8), cv2.IMREAD_COLOR)
     rgb = cv2.cvtColor(bgr, cv2.COLOR_BGR2RGB)
 
-    st.subheader(f"🔍 Preprocessing & Prediction for: {selected_filename}")
+    st.subheader("🔍 Preprocessing Steps")
     preprocessed = preprocess_with_steps(rgb)
     input_tensor = np.expand_dims(preprocessed, axis=0)
 
@@ -118,19 +127,15 @@ if uploaded_files and selected_filename:
     confidence = np.max(preds) * 100
 
     st.success(f"✅ Prediction: **{pred_label}** ({confidence:.2f}%)")
-
     show_lime(preprocessed, model, pred_idx, pred_label)
 
-# --- Show LIME for all images ---
+# -- Show LIME for all images with reduced size side-by-side --
 if uploaded_files:
     st.markdown("## 🧪 LIME Explanations for All Images")
     cols = st.columns(min(4, len(uploaded_files)))
     for i, file in enumerate(uploaded_files):
-        # Read bytes once and reset pointer
-        file_bytes = file.read()
         file.seek(0)
-
-        bgr = cv2.imdecode(np.frombuffer(file_bytes, np.uint8), cv2.IMREAD_COLOR)
+        bgr = cv2.imdecode(np.frombuffer(file.read(), np.uint8), cv2.IMREAD_COLOR)
         rgb = cv2.cvtColor(bgr, cv2.COLOR_BGR2RGB)
         img = cv2.resize(rgb, IMG_SIZE) / 255.0
         input_tensor = np.expand_dims(img, axis=0)
@@ -141,12 +146,19 @@ if uploaded_files:
 
         with cols[i % len(cols)]:
             st.markdown(f"**{file.name}**<br>🧠 *{pred_label}*", unsafe_allow_html=True)
+
             explanation = LIME_EXPLAINER.explain_instance(
                 image=img,
                 classifier_fn=lambda imgs: predict(imgs, model),
                 top_labels=1,
                 hide_color=0,
-                num_samples=1000
+                num_samples=50,
+            )
+            temp, mask = explanation.get_image_and_mask(
+                label=pred_idx, positive_only=True, num_features=10, hide_rest=False
             )
-            temp, mask = explanation.get_image_and_mask(label=pred_idx, positive_only=True, num_features=10, hide_rest=False)
-            st.image(mark_boundaries(temp, mask), use_column_width=True)
+            fig, ax = plt.subplots(figsize=(4, 3))
+            ax.imshow(mark_boundaries(temp, mask))
+            ax.axis("off")
+            st.pyplot(fig, use_container_width=False)
+            plt.close(fig)